Nikki Thean created KAFKA-4607:
----------------------------------
Summary: Kafka Streams allows you to provide strings with illegal
characters for internal topic names
Key: KAFKA-4607
URL: https://issues.apache.org/jira/browse/KAFKA-4607
Project: Kafka
Issue Type: Bug
Reporter: Nikki Thean
Priority: Minor
When using the aggregate function of the Kafka Streams DSL, I supplied the
function with a name for the underlying state store that contained whitespace.
The stream application ran with no error message but seemed to be hanging.
After a lot of troubleshooting, I looked in the logs on the Kafka cluster and
noticed that it was complaining about illegal characters in the log line. It
turns out that an internal changelog topic had been created that contained
whitespace and this was causing both Kafka and ZK to choke. I was unable to use
the delete-topic command line utility as it also threw the illegal character
error. In the end I had to manually delete the topics out of ZK and do some
stuff with the offset logs.
I am not well versed in the Kafka or Kafka Streams code, but from looking
through it, I notice that
[InternalTopicManager|https://github.com/apache/kafka/blob/trunk/streams/src/main/java/org/apache/kafka/streams/processor/internals/InternalTopicManager.java#L170]
does not seem to validate the names of auto-generated internal topics beyond
[checking for null topic
names|https://github.com/apache/kafka/blob/trunk/streams/src/main/java/org/apache/kafka/streams/processor/internals/InternalTopicConfig.java#L38]
rather than using Topic.validate.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
